Skip to content

Conversation

@bwbarrett
Copy link
Member

When running configure on the internal hwloc, enable plugins via
the --enable-plugins option when 1) --disable-dlopen was not
specified and 2) the user didn't add some variant of
--enable-plugins on the configure line already. The motivation
for this change is to not pull in hwloc dependencies like libcuda
in the general case.

Signed-off-by: Brian Barrett bbarrett@amazon.com

When running configure on the internal hwloc, enable plugins via
the --enable-plugins option when 1) --disable-dlopen was not
specified and 2) the user didn't add some variant of
--enable-plugins on the configure line already.  The motivation
for this change is to not pull in hwloc dependencies like libcuda
in the general case.

Signed-off-by: Brian Barrett <bbarrett@amazon.com>
@bwbarrett bwbarrett added this to the v5.0.0 milestone Jan 25, 2022
@bwbarrett bwbarrett requested a review from jsquyres January 25, 2022 18:06
@bwbarrett
Copy link
Member Author

This is in reference to the problem highlighted in #9907.

@bwbarrett
Copy link
Member Author

For posterity sake, a quick note about plugins from hwloc. In times past, that would cause some dependency problems, because hwloc was dlopened into the MPI executable by libopen-pal.so. With Open MPI 5.0 and master, hwloc is just a normal library that libopen-pal.so depends on, so there's no chained loading problems.

@bwbarrett bwbarrett merged commit 2328930 into open-mpi:master Jan 25, 2022
@bwbarrett bwbarrett deleted the bugfix/make-hwloc-build-cuda-plugin branch January 25, 2022 19:49
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants